SDB | SDGR | SGE | SMG | SMR | SRNA | SVD | SVGR | UVND | UVO | Main Page
NUTRIENT CONTROL OF BACTERIAL GENE EXPRESSION
| Michael Cashel, MD, PhD, Head, Section on Molecular Regulation
Jozsef Gal, PhD, Visiting Fellow |
|
|
Our goal is to understand how global patterns of bacterial gene expression are coordinated with nutrient availability. We continue to focus on the roles of (p)ppGpp, two regulatory nucleotide analogs related to GTP and GDP, but with a pyrophosphate esterified to the ribose 3'-hydroxyl. Nutrient limitation elevates (p)ppGpp, and nutrient sufficiency restores low basal levels, a mechanism that operates during starvation for amino acids, phosphate, nitrogen, or energy sources. Regulatory roles are assigned to (p)ppGpp; the elimination of (p)ppGpp can abolish regulation during starvation while the artificial elevation of (p)ppGpp (without starvation) mimics many regulatory effects of starvation. Responses to (p)ppGpp are a critical element in adaptive responses to starvation. We wish to understand the details of the molecular mechanisms whereby nutrient limitation governs (p)ppGpp metabolism and how regulation by (p)ppGpp works. Knowledge of these mechanisms is important for a basic understanding of the cell's interaction with nutrient availability. It is also of practical importance because an abundant literature links (p)ppGpp with bacterial pathogenicity in Gram-negative bacteria, synthesis of toxins and antibiotics in Gram-positives, and persistence of chronic infections in Mycobacteria tuberculosis, which occurs in the majority of the world's population. Discovery and structural documentation of an intrinsic protein switch for reciprocal control of (p)ppGpp synthesis and degradation Mechold,* Murphy, Cashel; in collaboration with Hilgenfeld, Hogg The balance of rates of (p)ppGpp synthesis and degradation determines (p)ppGpp's level as a nutritional stress signal. Regulation is achieved without simultaneously increasing both synthesis and degradation, which would be a futile waste of ATP. Activation of (p)ppGpp synthesis can be triggered when the supply of any aminoacyl tRNA fails to keep up with the demands of protein synthesis. An example is amino acid starvation leading to a decline in aminoacyl-tRNA and an increase in uncharged tRNA. Excess uncharged tRNA is then able to bind preferentially to ribosomal acceptor (A) sites dictated by now available "hungry" mRNA codons. We wish to understand the molecular details of the interaction between RelA, the ribosome, and the requirement for mRNA-specified uncharged tRNA binding. Last year, we reported finding that the Rsh protein from Streptococcus with its two catalytic domains in the N-terminal half of the protein (NTD) could exist in two activity states, depending on the presence of the C-terminal half-protein (CTD). In state 1 hydrolase is off, whereas synthetase is on; in state 2 hydrolase is on while synthetase is off. Hydrolase activity can change 50-fold, whereas synthetase activity varies about 15-fold. These effects are quantitatively consistent with findings made by other researchers using the full-length Rsh protein from M. tuberculosis on ribosomes bound with codon-specified uncharged tRNA. Our mutant mapping of hydrolase and synthetase catalytic domains found them to be nonoverlapping, verifiable with activity assays of peptides. We found unusual mis-sense mutants of one activity to be reactivated by reassembly with the missing domain, even though the CTD was absent. Subsequent isolation of domain-specific suppressors hinted that regions in one domain could regulate the activity of the other domain, independent of the CTD. We conclude that the switch leading to reciprocal regulation of opposing catalytic activities is an intrinsic "hard wired" feature of the NTD that could be changed by special NTD mutants, by NTD-CTD interactions, or by signals operating on the CTD. This year, an eight-year collaboration to determine the NTD structure has reached fruition by solving crystal structure at a resolution of 2.1 Å. The hydrolase structure closely resembles cyclic phosphodiesterases with a characteristic metal-dependent phosphohydrolase fold consisting of an up-down-up-down bundle of four alpha helices together with loop and a three-helix bundle. The synthetase structure, with its catalytic site 30 Å away from the hydrolase site, is a five-stranded mixed beta sheet sandwiched by five alpha helices. With the placement of two of the helices, the structure is similar to the nucleotidyl transferases superfamily, with the human DNA polymerase beta, polyA polymerases, and terminal deoxynucleotidyl transferases representing the closest structural examples. The structural relatedness is not evident from the linear amino acid sequence. The structure subjected to analysis consists of a dimer, with each monomer in a different conformation. Each monomeric conformation can be deduced to represent each of the two reciprocal activity states. We infer that allosteric transitions between the two monomeric conformations are attributable to ligand binding. Many unanswered questions remain regarding the predicted roles of individual residues. We also do not know how the regulatory CTD, uncharged tRNA, ribosomal binding, and environmental signals other than amino acid starvation provoke the transition between the two activity states.
Mechold U, Murphy H, Brown L, Cashel M. Intramolecular regulation of the opposing (p)ppGpp catalytic activities of RelSeq the Rel/Spo enzyme from Streptococcus equisimilis. J Bacteriol 2002;184:2878-888. Regulation of (p)ppGpp mediated by the SpoT protein Gal, Murphy, Cashel; in collaboration with Schneider We are also searching for mechanisms that lead to (p)ppGpp regulation through involvement of the SpoT protein, particularly with the CTD region that we know, from deletion studies, has regulatory functions. One approach involves characterization of spontaneous CTD region mutants found repeatedly among identical populations of E. coli B strains grown for over 20,000 generations by repeated overnight cultures in glucose minimal medium under standardized conditions. Researchers in Grenoble and at the NIH are characterizing these mutants. Typically, the new spoT CTD alleles appear to provide a 14 percent growth advantage over the wild type after a single overnight culture when introduced into the ancestral (generation 0) strain. In these evolving cultures, the spoT alleles appear to arise as a consequence of earlier spontaneous topA mutations, which also confer a growth advantage. It is unclear whether the growth advantage occurs at the level of steady-state growth or stationary-phase transitions. The onset of the stringent response to amino acid starvation appears to be accelerated in such mutants, but details remain to be characterized. Using a commercial (Bacteriomatch) bacterial two-hybrid system with the SpoT CTD as bait for screening an E. coli random fragment library, we have also searched for proteins interacting with SpoT protein. Our work led to the finding that protein fragments encoded by yrdA and fepE genes can bind to the SpoT CTD, verifiable by co-immunoprecipitation with a full-length SpoT-MalB fusion protein. The function of the yrdA gene is unknown, but its location is interesting. Its promoter is embedded in the upstream activation element of a very strong (rrnD) ribosomal RNA promoter region (rrnD) with transcripts differing from those of rRNA. This raises the possibility that abundance of yrdA transcripts might be reciprocally regulated by rrnD transcription, which is, in turn, known to be negatively regulated by ppGpp. We found that yrdA::lacZ fusion activities increase about 50-fold when rrnD transcripts are limited as a consequence of elevation of ppGpp. This finding led us to hypothesize that YrdA protein might act in a feedback regulatory circuit, with SpoT as a sensor of rRNA transcription. Induction of ppGpp is known to follow a damped harmonic pattern suggestive of such a circuit. Thus, high levels of ppGpp indirectly lead to overexpression of YrdA, which could promote its binding to SpoT CTD and then provoke lower ppGpp levels by either stimulating SpoT hydrolase or inhibiting SpoT synthetase. However, artificially overexpressing YrdA has few effects, including only a modest slowing of the turnoff of a rrnB-lacZ fusion during entry into stationary phase. This behavior weakly supports the hypothetical existence of the feedback loop. Deleting yrdA has no obvious phenotype even in indirect assays of (p)ppGpp levels. The role of FepE binding to SpoT is also uncertain. The E. coli K-12 fepE gene encodes a membrane protein homologous to an S. typhimurium protein that functions as a determinant of lipopolysaccharide O-antigen chain lengths. Long chain lengths are associated with increased virulence. Our strain of E. coli is defective in O-antigen synthesis because of an insertion element inactivation of the synthetic pathway; we are thus unable to verify this function of FepE. We have not found a dramatic phenotype accompanying either overexpression or deletion of the hydrophilic loop of FepE. Nevertheless, a function of FepE in E. coli analogous to S. typhimurium might explain of why deleting the spoT gene along with relA diminishes E. coli pathogenicity. Brown L, Gentry D, Elliott T, Cashel M. DksA affects ppGpp induction of RpoS at a translational level. J Bacteriol 2002;184:4455-4465. Cashel M, Hsu LM, Hernandez VJ. Changes in conserved region 3 of Escherichia coli sigma-70 re duce abortive transcription and enhance promoter escape. J Biol Chem 2003, 278:5539-5547. Murphy H, Cashel M. Isolation of RNA polymerase suppressors of a (p)ppGpp deficiency. In: Adhya S, Garges S, eds. Methods in Enzymology, vol. 371. New York: Academic Press, 2003; in press. COLLABORATORS Rolf Hilgenfeld, PhD, University of Lübeck, Germany
For further information, contact mcashel@nih.gov |
|
